Virtual router redundancy protocols (VRRPs) allow several routers on a multi-access link to utilize the same virtual internet protocol (IP) address. One example of a VRRP is described in the, Request for Comment (RFC) document series, document number RFC 2338. In accordance with the VRRP, one of the routers is selected as a master, while the other routers act as backup routers in the case of failure of the master router. Thus, several routers share a virtual IP address and media access control (MAC) address to form a “virtual routing system,” that consists of the several physical routers. Computers connected to the virtual routing system view the system as a single device, i.e., a signal router.
The particular physical router selected as the master at any given time performs routing of network packets through the virtual routing system. If the master router fails, however, one of the backup physical routers can be selected as a new master router so that routing through the virtual routing system can continue without significant interruption. VRRP eliminates single points of failure in network settings by providing a protocol that supports redundant router connections.
In many cases, VRRP allows for load sharing, in which different redundant routers are assigned as the master for different computers coupled to the virtual routing system. In that case, each redundant physical router of the virtual routing system carries traffic, but if one of the redundant physical routers fails, another of the redundant physical routers can be assigned the traffic of the failed router. In other words, when load sharing is used, different redundant physical routers operate as masters with respect to traffic received from different sources. Load sharing can improve throughput through the virtual routing system by utilizing the capabilities of each redundant physical router, while still providing the ability to adapt to failures of one of the physical routers.